👽

DoodlePix_pipeline.py

@@ -0,0 +1,977 @@
+import inspect
+from typing import Any, Callable, Dict, List, Optional, Union
+
+import numpy as np
+import PIL.Image
+import torch
+import torch.nn.functional as F
+from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer, CLIPVisionModelWithProjection
+import os
+
+from diffusers.callbacks import MultiPipelineCallbacks, PipelineCallback
+from diffusers.image_processor import PipelineImageInput, VaeImageProcessor
+from diffusers.loaders import FromSingleFileMixin, IPAdapterMixin, StableDiffusionLoraLoaderMixin, TextualInversionLoaderMixin
+from diffusers.models import AutoencoderKL, ImageProjection, UNet2DConditionModel
+from diffusers.schedulers import KarrasDiffusionSchedulers
+from diffusers.utils import PIL_INTERPOLATION, deprecate, is_torch_xla_available, logging
+from diffusers.utils.torch_utils import randn_tensor
+from diffusers.pipelines.pipeline_utils import DiffusionPipeline, StableDiffusionMixin
+from diffusers.pipelines.stable_diffusion import StableDiffusionPipelineOutput
+from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
+
+
+if is_torch_xla_available():
+    import torch_xla.core.xla_model as xm
+
+    XLA_AVAILABLE = True
+else:
+    XLA_AVAILABLE = False
+
+logger = logging.get_logger(__name__)  # pylint: disable=invalid-name
+
+
+# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img.preprocess
+def preprocess(image):
+    deprecation_message = "The preprocess method is deprecated and will be removed in diffusers 1.0.0. Please use VaeImageProcessor.preprocess(...) instead"
+    deprecate("preprocess", "1.0.0", deprecation_message, standard_warn=False)
+    if isinstance(image, torch.Tensor):
+        return image
+    elif isinstance(image, PIL.Image.Image):
+        image = [image]
+
+    if isinstance(image[0], PIL.Image.Image):
+        w, h = image[0].size
+        w, h = (x - x % 8 for x in (w, h))  # resize to integer multiple of 8
+
+        image = [np.array(i.resize((w, h), resample=PIL_INTERPOLATION["lanczos"]))[None, :] for i in image]
+        image = np.concatenate(image, axis=0)
+        image = np.array(image).astype(np.float32) / 255.0
+        image = image.transpose(0, 3, 1, 2)
+        image = 2.0 * image - 1.0
+        image = torch.from_numpy(image)
+    elif isinstance(image[0], torch.Tensor):
+        image = torch.cat(image, dim=0)
+    return image
+
+
+# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img.retrieve_latents
+def retrieve_latents(
+    encoder_output: torch.Tensor, generator: Optional[torch.Generator] = None, sample_mode: str = "sample"
+):
+    if hasattr(encoder_output, "latent_dist") and sample_mode == "sample":
+        return encoder_output.latent_dist.sample(generator)
+    elif hasattr(encoder_output, "latent_dist") and sample_mode == "argmax":
+        return encoder_output.latent_dist.mode()
+    elif hasattr(encoder_output, "latents"):
+        return encoder_output.latents
+    else:
+        raise AttributeError("Could not access latents of provided encoder_output")
+
+
+class StableDiffusionInstructPix2PixPipeline(
+    DiffusionPipeline,
+    StableDiffusionMixin,
+    TextualInversionLoaderMixin,
+    StableDiffusionLoraLoaderMixin,
+    IPAdapterMixin,
+    FromSingleFileMixin,
+):
+    r"""
+    Pipeline for pixel-level image editing by following text instructions (based on Stable Diffusion).
+
+    This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for the generic methods
+    implemented for all pipelines (downloading, saving, running on a particular device, etc.).
+
+    The pipeline also inherits the following loading methods:
+        - [`~loaders.TextualInversionLoaderMixin.load_textual_inversion`] for loading textual inversion embeddings
+        - [`~loaders.StableDiffusionLoraLoaderMixin.load_lora_weights`] for loading LoRA weights
+        - [`~loaders.StableDiffusionLoraLoaderMixin.save_lora_weights`] for saving LoRA weights
+        - [`~loaders.IPAdapterMixin.load_ip_adapter`] for loading IP Adapters
+
+    Args:
+        vae ([`AutoencoderKL`]):
+            Variational Auto-Encoder (VAE) model to encode and decode images to and from latent representations.
+        text_encoder ([`~transformers.CLIPTextModel`]):
+            Frozen text-encoder ([clip-vit-large-patch14](https://huggingface.co/openai/clip-vit-large-patch14)).
+        tokenizer ([`~transformers.CLIPTokenizer`]):
+            A `CLIPTokenizer` to tokenize text.
+        unet ([`UNet2DConditionModel`]):
+            A `UNet2DConditionModel` to denoise the encoded image latents.
+        scheduler ([`SchedulerMixin`]):
+            A scheduler to be used in combination with `unet` to denoise the encoded image latents. Can be one of
+            [`DDIMScheduler`], [`LMSDiscreteScheduler`], or [`PNDMScheduler`].
+        safety_checker ([`StableDiffusionSafetyChecker`]):
+            Classification module that estimates whether generated images could be considered offensive or harmful.
+            Please refer to the [model card](https://huggingface.co/stable-diffusion-v1-5/stable-diffusion-v1-5) for
+            more details about a model's potential harms.
+        feature_extractor ([`~transformers.CLIPImageProcessor`]):
+            A `CLIPImageProcessor` to extract features from generated images; used as inputs to the `safety_checker`.
+    """
+
+    model_cpu_offload_seq = "text_encoder->unet->vae"
+    _optional_components = ["safety_checker", "feature_extractor", "image_encoder"]
+    _exclude_from_cpu_offload = ["safety_checker"]
+    _callback_tensor_inputs = ["latents", "prompt_embeds", "image_latents"]
+
+    def __init__(
+        self,
+        vae: AutoencoderKL,
+        text_encoder: CLIPTextModel,
+        tokenizer: CLIPTokenizer,
+        unet: UNet2DConditionModel,
+        scheduler: KarrasDiffusionSchedulers,
+        safety_checker: StableDiffusionSafetyChecker,
+        feature_extractor: CLIPImageProcessor,
+        image_encoder: Optional[CLIPVisionModelWithProjection] = None,
+        requires_safety_checker: bool = False,
+        teacher_text_encoder: Optional[CLIPTextModel] = None,
+        teacher_loss_weight: float = 0.1,
+        fidelity_mlp = None,
+    ):
+        super().__init__()
+
+        if safety_checker is None and requires_safety_checker:
+            logger.warning(
+                f"You have disabled the safety checker for {self.__class__} by passing `safety_checker=None`. Ensure"
+                " that you abide to the conditions of the Stable Diffusion license and do not expose unfiltered"
+                " results in services or applications open to the public. Both the diffusers team and Hugging Face"
+                " strongly recommend to keep the safety filter enabled in all public facing circumstances, disabling"
+                " it only for use-cases that involve analyzing network behavior or auditing its results. For more"
+                " information, please have a look at https://github.com/huggingface/diffusers/pull/254 ."
+            )
+
+        if safety_checker is not None and feature_extractor is None:
+            raise ValueError(
+                "Make sure to define a feature extractor when loading {self.__class__} if you want to use the safety"
+                " checker. If you do not want to use the safety checker, you can pass `'safety_checker=None'` instead."
+            )
+
+        self.register_modules(
+            vae=vae,
+            text_encoder=text_encoder,
+            tokenizer=tokenizer,
+            unet=unet,
+            scheduler=scheduler,
+            safety_checker=safety_checker,
+            feature_extractor=feature_extractor,
+            image_encoder=image_encoder,
+        )
+        self.teacher_text_encoder = teacher_text_encoder
+        self.teacher_loss_weight = teacher_loss_weight
+        self.fidelity_mlp = fidelity_mlp
+        if self.teacher_text_encoder is not None:
+            self.teacher_text_encoder.eval()
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8
+        self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor)
+        self.register_to_config(requires_safety_checker=requires_safety_checker)
+
+    @torch.no_grad()
+    def __call__(
+        self,
+        prompt: Union[str, List[str]] = None,
+        image: PipelineImageInput = None,
+        num_inference_steps: int = 100,
+        guidance_scale: float = 7.5,
+        image_guidance_scale: float = 1.5,
+        negative_prompt: Optional[Union[str, List[str]]] = None,
+        num_images_per_prompt: Optional[int] = 1,
+        eta: float = 0.0,
+        generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None,
+        latents: Optional[torch.Tensor] = None,
+        prompt_embeds: Optional[torch.Tensor] = None,
+        negative_prompt_embeds: Optional[torch.Tensor] = None,
+        ip_adapter_image: Optional[PipelineImageInput] = None,
+        ip_adapter_image_embeds: Optional[List[torch.Tensor]] = None,
+        output_type: Optional[str] = "pil",
+        return_dict: bool = True,
+        callback_on_step_end: Optional[
+            Union[Callable[[int, int, Dict], None], PipelineCallback, MultiPipelineCallbacks]
+        ] = None,
+        callback_on_step_end_tensor_inputs: List[str] = ["latents"],
+        cross_attention_kwargs: Optional[Dict[str, Any]] = None,
+        **kwargs,
+    ):
+        r"""
+        The call function to the pipeline for generation.
+
+        Args:
+            prompt (`str` or `List[str]`, *optional*):
+                The prompt or prompts to guide image generation. If not defined, you need to pass `prompt_embeds`.
+            image (`torch.Tensor` `np.ndarray`, `PIL.Image.Image`, `List[torch.Tensor]`, `List[PIL.Image.Image]`, or `List[np.ndarray]`):
+                `Image` or tensor representing an image batch to be repainted according to `prompt`. Can also accept
+                image latents as `image`, but if passing latents directly it is not encoded again.
+            num_inference_steps (`int`, *optional*, defaults to 100):
+                The number of denoising steps. More denoising steps usually lead to a higher quality image at the
+                expense of slower inference.
+            guidance_scale (`float`, *optional*, defaults to 7.5):
+                A higher guidance scale value encourages the model to generate images closely linked to the text
+                `prompt` at the expense of lower image quality. Guidance scale is enabled when `guidance_scale > 1`.
+            image_guidance_scale (`float`, *optional*, defaults to 1.5):
+                Push the generated image towards the initial `image`. Image guidance scale is enabled by setting
+                `image_guidance_scale > 1`. Higher image guidance scale encourages generated images that are closely
+                linked to the source `image`, usually at the expense of lower image quality. This pipeline requires a
+                value of at least `1`.
+            negative_prompt (`str` or `List[str]`, *optional*):
+                The prompt or prompts to guide what to not include in image generation. If not defined, you need to
+                pass `negative_prompt_embeds` instead. Ignored when not using guidance (`guidance_scale < 1`).
+            num_images_per_prompt (`int`, *optional*, defaults to 1):
+                The number of images to generate per prompt.
+            eta (`float`, *optional*, defaults to 0.0):
+                Corresponds to parameter eta (η) from the [DDIM](https://arxiv.org/abs/2010.02502) paper. Only applies
+                to the [`~schedulers.DDIMScheduler`], and is ignored in other schedulers.
+            generator (`torch.Generator`, *optional*):
+                A [`torch.Generator`](https://pytorch.org/docs/stable/generated/torch.Generator.html) to make
+                generation deterministic.
+            latents (`torch.Tensor`, *optional*):
+                Pre-generated noisy latents sampled from a Gaussian distribution, to be used as inputs for image
+                generation. Can be used to tweak the same generation with different prompts. If not provided, a latents
+                tensor is generated by sampling using the supplied random `generator`.
+            prompt_embeds (`torch.Tensor`, *optional*):
+                Pre-generated text embeddings. Can be used to easily tweak text inputs (prompt weighting). If not
+                provided, text embeddings are generated from the `prompt` input argument.
+            negative_prompt_embeds (`torch.Tensor`, *optional*):
+                Pre-generated negative text embeddings. Can be used to easily tweak text inputs (prompt weighting). If
+                not provided, `negative_prompt_embeds` are generated from the `negative_prompt` input argument.
+            ip_adapter_image: (`PipelineImageInput`, *optional*):
+                Optional image input to work with IP Adapters.
+            output_type (`str`, *optional*, defaults to `"pil"`):
+                The output format of the generated image. Choose between `PIL.Image` or `np.array`.
+            return_dict (`bool`, *optional*, defaults to `True`):
+                Whether or not to return a [`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] instead of a
+                plain tuple.
+            callback_on_step_end (`Callable`, `PipelineCallback`, `MultiPipelineCallbacks`, *optional*):
+                A function or a subclass of `PipelineCallback` or `MultiPipelineCallbacks` that is called at the end of
+                each denoising step during the inference. with the following arguments: `callback_on_step_end(self:
+                DiffusionPipeline, step: int, timestep: int, callback_kwargs: Dict)`. `callback_kwargs` will include a
+                list of all tensors as specified by `callback_on_step_end_tensor_inputs`.
+            callback_on_step_end_tensor_inputs (`List`, *optional*):
+                The list of tensor inputs for the `callback_on_step_end` function. The tensors specified in the list
+                will be passed as `callback_kwargs` argument. You will only be able to include variables listed in the
+                `._callback_tensor_inputs` attribute of your pipeline class.
+            cross_attention_kwargs (`dict`, *optional*):
+                A kwargs dictionary that if specified is passed along to the [`AttentionProcessor`] as defined in
+                [`self.processor`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py).
+
+        Examples:
+
+        ```py
+        >>> import PIL
+        >>> import requests
+        >>> import torch
+        >>> from io import BytesIO
+
+        >>> from diffusers import StableDiffusionInstructPix2PixPipeline
+
+
+        >>> def download_image(url):
+        ...     response = requests.get(url)
+        ...     return PIL.Image.open(BytesIO(response.content)).convert("RGB")
+
+
+        >>> img_url = "https://huggingface.co/datasets/diffusers/diffusers-images-docs/resolve/main/mountain.png"
+
+        >>> image = download_image(img_url).resize((512, 512))
+
+        >>> pipe = StableDiffusionInstructPix2PixPipeline.from_pretrained(
+        ...     "timbrooks/instruct-pix2pix", torch_dtype=torch.float16
+        ... )
+        >>> pipe = pipe.to("cuda")
+
+        >>> prompt = "make the mountains snowy"
+        >>> image = pipe(prompt=prompt, image=image).images[0]
+        ```
+
+        Returns:
+            [`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] or `tuple`:
+                If `return_dict` is `True`, [`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] is returned,
+                otherwise a `tuple` is returned where the first element is a list with the generated images and the
+                second element is a list of `bool`s indicating whether the corresponding generated image contains
+                "not-safe-for-work" (nsfw) content.
+        """
+
+        callback = kwargs.pop("callback", None)
+        callback_steps = kwargs.pop("callback_steps", None)
+
+        if callback is not None:
+            deprecate(
+                "callback",
+                "1.0.0",
+                "Passing `callback` as an input argument to `__call__` is deprecated, consider use `callback_on_step_end`",
+            )
+        if callback_steps is not None:
+            deprecate(
+                "callback_steps",
+                "1.0.0",
+                "Passing `callback_steps` as an input argument to `__call__` is deprecated, consider use `callback_on_step_end`",
+            )
+
+        if isinstance(callback_on_step_end, (PipelineCallback, MultiPipelineCallbacks)):
+            callback_on_step_end_tensor_inputs = callback_on_step_end.tensor_inputs
+
+        # 0. Check inputs
+        self.check_inputs(
+            prompt,
+            callback_steps,
+            negative_prompt,
+            prompt_embeds,
+            negative_prompt_embeds,
+            ip_adapter_image,
+            ip_adapter_image_embeds,
+            callback_on_step_end_tensor_inputs,
+        )
+        self._guidance_scale = guidance_scale
+        self._image_guidance_scale = image_guidance_scale
+
+        device = self._execution_device
+
+        if image is None:
+            raise ValueError("`image` input cannot be undefined.")
+
+        # 1. Define call parameters
+        if prompt is not None and isinstance(prompt, str):
+            batch_size = 1
+        elif prompt is not None and isinstance(prompt, list):
+            batch_size = len(prompt)
+        else:
+            batch_size = prompt_embeds.shape[0]
+
+        device = self._execution_device
+
+        # 2. Encode input prompt
+        prompt_embeds = self._encode_prompt(
+            prompt,
+            device,
+            num_images_per_prompt,
+            self.do_classifier_free_guidance,
+            negative_prompt,
+            prompt_embeds=prompt_embeds,
+            negative_prompt_embeds=negative_prompt_embeds,
+            return_teacher_loss=False,
+        )
+
+        if ip_adapter_image is not None or ip_adapter_image_embeds is not None:
+            image_embeds = self.prepare_ip_adapter_image_embeds(
+                ip_adapter_image,
+                ip_adapter_image_embeds,
+                device,
+                batch_size * num_images_per_prompt,
+                self.do_classifier_free_guidance,
+            )
+        # 3. Preprocess image
+        image = self.image_processor.preprocess(image)
+
+        # 4. set timesteps
+        self.scheduler.set_timesteps(num_inference_steps, device=device)
+        timesteps = self.scheduler.timesteps
+
+        # 5. Prepare Image latents
+        image_latents = self.prepare_image_latents(
+            image,
+            batch_size,
+            num_images_per_prompt,
+            prompt_embeds.dtype,
+            device,
+            self.do_classifier_free_guidance,
+        )
+
+        height, width = image_latents.shape[-2:]
+        height = height * self.vae_scale_factor
+        width = width * self.vae_scale_factor
+
+        # 6. Prepare latent variables
+        num_channels_latents = self.vae.config.latent_channels
+        latents = self.prepare_latents(
+            batch_size * num_images_per_prompt,
+            num_channels_latents,
+            height,
+            width,
+            prompt_embeds.dtype,
+            device,
+            generator,
+            latents,
+        )
+
+        # 7. Check that shapes of latents and image match the UNet channels
+        num_channels_image = image_latents.shape[1]
+        if num_channels_latents + num_channels_image != self.unet.config.in_channels:
+            raise ValueError(
+                f"Incorrect configuration settings! The config of `pipeline.unet`: {self.unet.config} expects"
+                f" {self.unet.config.in_channels} but received `num_channels_latents`: {num_channels_latents} +"
+                f" `num_channels_image`: {num_channels_image} "
+                f" = {num_channels_latents+num_channels_image}. Please verify the config of"
+                " `pipeline.unet` or your `image` input."
+            )
+
+        # 8. Prepare extra step kwargs. TODO: Logic should ideally just be moved out of the pipeline
+        extra_step_kwargs = self.prepare_extra_step_kwargs(generator, eta)
+
+        # 8.1 Add image embeds for IP-Adapter
+        added_cond_kwargs = {"image_embeds": image_embeds} if ip_adapter_image is not None else None
+
+        # 9. Denoising loop
+        num_warmup_steps = len(timesteps) - num_inference_steps * self.scheduler.order
+        self._num_timesteps = len(timesteps)
+        with self.progress_bar(total=num_inference_steps) as progress_bar:
+            for i, t in enumerate(timesteps):
+                # Expand the latents if we are doing classifier free guidance.
+                # The latents are expanded 3 times because for pix2pix the guidance\
+                # is applied for both the text and the input image.
+                latent_model_input = torch.cat([latents] * 3) if self.do_classifier_free_guidance else latents
+
+                # concat latents, image_latents in the channel dimension
+                scaled_latent_model_input = self.scheduler.scale_model_input(latent_model_input, t)
+                scaled_latent_model_input = torch.cat([scaled_latent_model_input, image_latents], dim=1)
+
+                # predict the noise residual
+                noise_pred = self.unet(
+                    scaled_latent_model_input,
+                    t,
+                    encoder_hidden_states=prompt_embeds,
+                    added_cond_kwargs=added_cond_kwargs,
+                    cross_attention_kwargs=cross_attention_kwargs,
+                    return_dict=False,
+                )[0]
+
+                # perform guidance
+                if self.do_classifier_free_guidance:
+                    noise_pred_text, noise_pred_image, noise_pred_uncond = noise_pred.chunk(3)
+                    noise_pred = (
+                        noise_pred_uncond
+                        + self.guidance_scale * (noise_pred_text - noise_pred_image)
+                        + self.image_guidance_scale * (noise_pred_image - noise_pred_uncond)
+                    )
+
+                # compute the previous noisy sample x_t -> x_t-1
+                latents = self.scheduler.step(noise_pred, t, latents, **extra_step_kwargs, return_dict=False)[0]
+
+                if callback_on_step_end is not None:
+                    callback_kwargs = {}
+                    for k in callback_on_step_end_tensor_inputs:
+                        callback_kwargs[k] = locals()[k]
+                    callback_outputs = callback_on_step_end(self, i, t, callback_kwargs)
+
+                    latents = callback_outputs.pop("latents", latents)
+                    prompt_embeds = callback_outputs.pop("prompt_embeds", prompt_embeds)
+                    negative_prompt_embeds = callback_outputs.pop("negative_prompt_embeds", negative_prompt_embeds)
+                    image_latents = callback_outputs.pop("image_latents", image_latents)
+
+                # call the callback, if provided
+                if i == len(timesteps) - 1 or ((i + 1) > num_warmup_steps and (i + 1) % self.scheduler.order == 0):
+                    progress_bar.update()
+                    if callback is not None and i % callback_steps == 0:
+                        step_idx = i // getattr(self.scheduler, "order", 1)
+                        callback(step_idx, t, latents)
+
+                if XLA_AVAILABLE:
+                    xm.mark_step()
+
+        if not output_type == "latent":
+            image = self.vae.decode(latents / self.vae.config.scaling_factor, return_dict=False)[0]
+            image, has_nsfw_concept = self.run_safety_checker(image, device, prompt_embeds.dtype)
+        else:
+            image = latents
+            has_nsfw_concept = None
+
+        if has_nsfw_concept is None:
+            do_denormalize = [True] * image.shape[0]
+        else:
+            do_denormalize = [not has_nsfw for has_nsfw in has_nsfw_concept]
+
+        image = self.image_processor.postprocess(image, output_type=output_type, do_denormalize=do_denormalize)
+
+        # Offload all models
+        self.maybe_free_model_hooks()
+
+        if not return_dict:
+            return (image, has_nsfw_concept)
+
+        return StableDiffusionPipelineOutput(images=image, nsfw_content_detected=has_nsfw_concept)
+
+    def _encode_prompt(
+        self,
+        prompt,
+        device,
+        num_images_per_prompt,
+        do_classifier_free_guidance,
+        negative_prompt=None,
+        prompt_embeds: Optional[torch.Tensor] = None,
+        negative_prompt_embeds: Optional[torch.Tensor] = None,
+        return_teacher_loss: bool = False,
+    ):
+        r"""
+        Encodes the prompt into text encoder hidden states.
+
+        Args:
+             prompt (`str` or `List[str]`, *optional*):
+                prompt to be encoded
+            device: (`torch.device`):
+                torch device
+            num_images_per_prompt (`int`):
+                number of images that should be generated per prompt
+            do_classifier_free_guidance (`bool`):
+                whether to use classifier free guidance or not
+            negative_ prompt (`str` or `List[str]`, *optional*):
+                The prompt or prompts not to guide the image generation. If not defined, one has to pass
+                `negative_prompt_embeds` instead. Ignored when not using guidance (i.e., ignored if `guidance_scale` is
+                less than `1`).
+            prompt_embeds (`torch.Tensor`, *optional*):
+                Pre-generated text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt weighting. If not
+                provided, text embeddings will be generated from `prompt` input argument.
+            negative_prompt_embeds (`torch.Tensor`, *optional*):
+                Pre-generated negative text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt
+                weighting. If not provided, negative_prompt_embeds will be generated from `negative_prompt` input
+                argument.
+        """
+        if prompt is not None and isinstance(prompt, str):
+            batch_size = 1
+        elif prompt is not None and isinstance(prompt, list):
+            batch_size = len(prompt)
+        else:
+            batch_size = prompt_embeds.shape[0]
+
+        if prompt_embeds is None:
+            # textual inversion: process multi-vector tokens if necessary
+            if isinstance(self, TextualInversionLoaderMixin):
+                prompt = self.maybe_convert_prompt(prompt, self.tokenizer)
+
+            text_inputs = self.tokenizer(
+                prompt,
+                padding="max_length",
+                max_length=self.tokenizer.model_max_length,
+                truncation=True,
+                return_tensors="pt",
+            )
+            text_input_ids = text_inputs.input_ids
+            untruncated_ids = self.tokenizer(prompt, padding="longest", return_tensors="pt").input_ids
+
+            if untruncated_ids.shape[-1] >= text_input_ids.shape[-1] and not torch.equal(
+                text_input_ids, untruncated_ids
+            ):
+                removed_text = self.tokenizer.batch_decode(
+                    untruncated_ids[:, self.tokenizer.model_max_length - 1 : -1]
+                )
+                logger.warning(
+                    "The following part of your input was truncated because CLIP can only handle sequences up to"
+                    f" {self.tokenizer.model_max_length} tokens: {removed_text}"
+                )
+
+            if hasattr(self.text_encoder.config, "use_attention_mask") and self.text_encoder.config.use_attention_mask:
+                attention_mask = text_inputs.attention_mask.to(device)
+            else:
+                attention_mask = None
+
+            prompt_embeds = self.text_encoder(text_input_ids.to(device), attention_mask=attention_mask)
+            prompt_embeds = prompt_embeds[0]
+
+        prompt_embeds = prompt_embeds.to(dtype=self.text_encoder.dtype, device=device)
+
+        # Extract fidelity value from the prompt
+        fidelity_val = 0.5  # default fidelity
+        if self.fidelity_mlp is not None:
+            if prompt is not None:
+                if isinstance(prompt, str):
+                    import re
+                    match = re.search(r"f\s*=?\s*(\d+)|f(\d+)", prompt, re.IGNORECASE)
+                    if match:
+                        f_int = int(match.group(1) if match.group(1) else match.group(2))
+                        f_int = max(0, min(f_int, 9))
+                        fidelity_val = f_int / 9.0
+                elif isinstance(prompt, list):
+                    import re
+                    f_vals = []
+                    for p in prompt:
+                        match = re.search(r"f\s*=?\s*(\d+)|f(\d+)", p, re.IGNORECASE)
+                        if match:
+                            f_int = int(match.group(1) if match.group(1) else match.group(2))
+                            f_int = max(0, min(f_int, 9))
+                            f_vals.append(f_int / 9.0)
+                    if f_vals:
+                        fidelity_val = sum(f_vals) / len(f_vals)
+
+            # Create fidelity tensor
+            batch_size = prompt_embeds.shape[0]
+            fidelity_tensor = torch.full((batch_size, 1), fidelity_val, device=device, dtype=prompt_embeds.dtype)
+            
+            # Get fidelity embedding
+            fidelity_embedding = self.fidelity_mlp(fidelity_tensor)  # (batch, hidden_size)
+            
+            # This keeps the sequence length the same (77 tokens)
+            prompt_embeds[:, 0] = prompt_embeds[:, 0] + (0.8 * fidelity_embedding)
+
+        bs_embed, seq_len, _ = prompt_embeds.shape
+        # duplicate text embeddings for each generation per prompt
+        prompt_embeds = prompt_embeds.repeat(1, num_images_per_prompt, 1)
+        prompt_embeds = prompt_embeds.view(bs_embed * num_images_per_prompt, seq_len, -1)
+
+        teacher_loss = None
+        if return_teacher_loss and self.teacher_text_encoder is not None:
+            # Compute teacher embeddings using the frozen teacher text encoder
+            with torch.no_grad():
+                teacher_outputs = self.teacher_text_encoder(text_input_ids.to(device), attention_mask=attention_mask)
+                teacher_embeds = teacher_outputs[0]
+            teacher_embeds = teacher_embeds.to(dtype=self.text_encoder.dtype, device=device)
+            teacher_embeds = teacher_embeds.repeat(1, num_images_per_prompt, 1)
+            teacher_embeds = teacher_embeds.view(bs_embed * num_images_per_prompt, seq_len, -1)
+            # Compute an MSE loss between teacher and student text embeddings
+            teacher_loss = F.mse_loss(prompt_embeds, teacher_embeds) * self.teacher_loss_weight
+
+        if do_classifier_free_guidance:
+            # get unconditional embeddings for classifier free guidance
+            if negative_prompt_embeds is None:
+                uncond_tokens: List[str]
+                if negative_prompt is None:
+                    uncond_tokens = [""] * batch_size
+                elif isinstance(negative_prompt, str):
+                    uncond_tokens = [negative_prompt]
+                elif batch_size != len(negative_prompt):
+                    raise ValueError(
+                        f"`negative_prompt`: {negative_prompt} has batch size {len(negative_prompt)}, but `prompt`:"
+                        f" {prompt} has batch size {batch_size}. Please make sure that passed `negative_prompt` matches"
+                        " the batch size of `prompt`."
+                    )
+                else:
+                    uncond_tokens = negative_prompt
+
+                if isinstance(self, TextualInversionLoaderMixin):
+                    uncond_tokens = self.maybe_convert_prompt(uncond_tokens, self.tokenizer)
+
+                max_length = prompt_embeds.shape[1]
+                uncond_input = self.tokenizer(
+                    uncond_tokens,
+                    padding="max_length",
+                    max_length=max_length,
+                    truncation=True,
+                    return_tensors="pt",
+                )
+
+                if hasattr(self.text_encoder.config, "use_attention_mask") and self.text_encoder.config.use_attention_mask:
+                    attention_mask = uncond_input.attention_mask.to(device)
+                else:
+                    attention_mask = None
+
+                negative_prompt_embeds = self.text_encoder(
+                    uncond_input.input_ids.to(device),
+                    attention_mask=attention_mask,
+                )[0]
+
+            negative_prompt_embeds = negative_prompt_embeds.to(dtype=self.text_encoder.dtype, device=device)
+            
+            # Apply the same fidelity modification to negative prompt embeddings
+            if self.fidelity_mlp is not None:
+                bs_neg = negative_prompt_embeds.shape[0]
+                fidelity_tensor = torch.full((bs_neg, 1), fidelity_val, device=device, dtype=negative_prompt_embeds.dtype)
+                fidelity_embedding = self.fidelity_mlp(fidelity_tensor)
+                
+                # FIXED: Modify the first token of negative embedding instead of concatenating
+                negative_prompt_embeds[:, 0] = negative_prompt_embeds[:, 0] + 0.2 * fidelity_embedding
+            
+            bs_embed, seq_len, _ = negative_prompt_embeds.shape
+            negative_prompt_embeds = negative_prompt_embeds.repeat(1, num_images_per_prompt, 1)
+            negative_prompt_embeds = negative_prompt_embeds.view(batch_size * num_images_per_prompt, seq_len, -1)
+
+            # For classifier free guidance, we need to do it for both positive and negative
+            prompt_embeds = torch.cat([prompt_embeds, negative_prompt_embeds, negative_prompt_embeds])
+
+        if return_teacher_loss:
+            return prompt_embeds, teacher_loss
+        return prompt_embeds
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.encode_image
+    def encode_image(self, image, device, num_images_per_prompt, output_hidden_states=None):
+        dtype = next(self.image_encoder.parameters()).dtype
+
+        if not isinstance(image, torch.Tensor):
+            image = self.feature_extractor(image, return_tensors="pt").pixel_values
+
+        image = image.to(device=device, dtype=dtype)
+        if output_hidden_states:
+            image_enc_hidden_states = self.image_encoder(image, output_hidden_states=True).hidden_states[-2]
+            image_enc_hidden_states = image_enc_hidden_states.repeat_interleave(num_images_per_prompt, dim=0)
+            uncond_image_enc_hidden_states = self.image_encoder(
+                torch.zeros_like(image), output_hidden_states=True
+            ).hidden_states[-2]
+            uncond_image_enc_hidden_states = uncond_image_enc_hidden_states.repeat_interleave(
+                num_images_per_prompt, dim=0
+            )
+            return image_enc_hidden_states, uncond_image_enc_hidden_states
+        else:
+            image_embeds = self.image_encoder(image).image_embeds
+            image_embeds = image_embeds.repeat_interleave(num_images_per_prompt, dim=0)
+            uncond_image_embeds = torch.zeros_like(image_embeds)
+
+            return image_embeds, uncond_image_embeds
+
+    def prepare_ip_adapter_image_embeds(
+        self, ip_adapter_image, ip_adapter_image_embeds, device, num_images_per_prompt, do_classifier_free_guidance
+    ):
+        if ip_adapter_image_embeds is None:
+            if not isinstance(ip_adapter_image, list):
+                ip_adapter_image = [ip_adapter_image]
+
+            if len(ip_adapter_image) != len(self.unet.encoder_hid_proj.image_projection_layers):
+                raise ValueError(
+                    f"`ip_adapter_image` must have same length as the number of IP Adapters. Got {len(ip_adapter_image)} images and {len(self.unet.encoder_hid_proj.image_projection_layers)} IP Adapters."
+                )
+
+            image_embeds = []
+            for single_ip_adapter_image, image_proj_layer in zip(
+                ip_adapter_image, self.unet.encoder_hid_proj.image_projection_layers
+            ):
+                output_hidden_state = not isinstance(image_proj_layer, ImageProjection)
+                single_image_embeds, single_negative_image_embeds = self.encode_image(
+                    single_ip_adapter_image, device, 1, output_hidden_state
+                )
+                single_image_embeds = torch.stack([single_image_embeds] * num_images_per_prompt, dim=0)
+                single_negative_image_embeds = torch.stack(
+                    [single_negative_image_embeds] * num_images_per_prompt, dim=0
+                )
+
+                if do_classifier_free_guidance:
+                    single_image_embeds = torch.cat(
+                        [single_image_embeds, single_negative_image_embeds, single_negative_image_embeds]
+                    )
+                    single_image_embeds = single_image_embeds.to(device)
+
+                image_embeds.append(single_image_embeds)
+        else:
+            repeat_dims = [1]
+            image_embeds = []
+            for single_image_embeds in ip_adapter_image_embeds:
+                if do_classifier_free_guidance:
+                    (
+                        single_image_embeds,
+                        single_negative_image_embeds,
+                        single_negative_image_embeds,
+                    ) = single_image_embeds.chunk(3)
+                    single_image_embeds = single_image_embeds.repeat(
+                        num_images_per_prompt, *(repeat_dims * len(single_image_embeds.shape[1:]))
+                    )
+                    single_negative_image_embeds = single_negative_image_embeds.repeat(
+                        num_images_per_prompt, *(repeat_dims * len(single_negative_image_embeds.shape[1:]))
+                    )
+                    single_image_embeds = torch.cat(
+                        [single_image_embeds, single_negative_image_embeds, single_negative_image_embeds]
+                    )
+                else:
+                    single_image_embeds = single_image_embeds.repeat(
+                        num_images_per_prompt, *(repeat_dims * len(single_image_embeds.shape[1:]))
+                    )
+                image_embeds.append(single_image_embeds)
+
+        return image_embeds
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.run_safety_checker
+    def run_safety_checker(self, image, device, dtype):
+        if self.safety_checker is None:
+            has_nsfw_concept = None
+        else:
+            if torch.is_tensor(image):
+                feature_extractor_input = self.image_processor.postprocess(image, output_type="pil")
+            else:
+                feature_extractor_input = self.image_processor.numpy_to_pil(image)
+            safety_checker_input = self.feature_extractor(feature_extractor_input, return_tensors="pt").to(device)
+            image, has_nsfw_concept = self.safety_checker(
+                images=image, clip_input=safety_checker_input.pixel_values.to(dtype)
+            )
+        return image, has_nsfw_concept
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_extra_step_kwargs
+    def prepare_extra_step_kwargs(self, generator, eta):
+        # prepare extra kwargs for the scheduler step, since not all schedulers have the same signature
+        # eta (η) is only used with the DDIMScheduler, it will be ignored for other schedulers.
+        # eta corresponds to η in DDIM paper: https://arxiv.org/abs/2010.02502
+        # and should be between [0, 1]
+
+        accepts_eta = "eta" in set(inspect.signature(self.scheduler.step).parameters.keys())
+        extra_step_kwargs = {}
+        if accepts_eta:
+            extra_step_kwargs["eta"] = eta
+
+        # check if the scheduler accepts generator
+        accepts_generator = "generator" in set(inspect.signature(self.scheduler.step).parameters.keys())
+        if accepts_generator:
+            extra_step_kwargs["generator"] = generator
+        return extra_step_kwargs
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.decode_latents
+    def decode_latents(self, latents):
+        deprecation_message = "The decode_latents method is deprecated and will be removed in 1.0.0. Please use VaeImageProcessor.postprocess(...) instead"
+        deprecate("decode_latents", "1.0.0", deprecation_message, standard_warn=False)
+
+        latents = 1 / self.vae.config.scaling_factor * latents
+        image = self.vae.decode(latents, return_dict=False)[0]
+        image = (image / 2 + 0.5).clamp(0, 1)
+        # we always cast to float32 as this does not cause significant overhead and is compatible with bfloat16
+        image = image.cpu().permute(0, 2, 3, 1).float().numpy()
+        return image
+
+    def check_inputs(
+        self,
+        prompt,
+        callback_steps,
+        negative_prompt=None,
+        prompt_embeds=None,
+        negative_prompt_embeds=None,
+        ip_adapter_image=None,
+        ip_adapter_image_embeds=None,
+        callback_on_step_end_tensor_inputs=None,
+    ):
+        if callback_steps is not None and (not isinstance(callback_steps, int) or callback_steps <= 0):
+            raise ValueError(
+                f"`callback_steps` has to be a positive integer but is {callback_steps} of type"
+                f" {type(callback_steps)}."
+            )
+
+        if callback_on_step_end_tensor_inputs is not None and not all(
+            k in self._callback_tensor_inputs for k in callback_on_step_end_tensor_inputs
+        ):
+            raise ValueError(
+                f"`callback_on_step_end_tensor_inputs` has to be in {self._callback_tensor_inputs}, but found {[k for k in callback_on_step_end_tensor_inputs if k not in self._callback_tensor_inputs]}"
+            )
+
+        if prompt is not None and prompt_embeds is not None:
+            raise ValueError(
+                f"Cannot forward both `prompt`: {prompt} and `prompt_embeds`: {prompt_embeds}. Please make sure to"
+                " only forward one of the two."
+            )
+        elif prompt is None and prompt_embeds is None:
+            raise ValueError(
+                "Provide either `prompt` or `prompt_embeds`. Cannot leave both `prompt` and `prompt_embeds` undefined."
+            )
+        elif prompt is not None and (not isinstance(prompt, str) and not isinstance(prompt, list)):
+            raise ValueError(f"`prompt` has to be of type `str` or `list` but is {type(prompt)}")
+
+        if negative_prompt is not None and negative_prompt_embeds is not None:
+            raise ValueError(
+                f"Cannot forward both `negative_prompt`: {negative_prompt} and `negative_prompt_embeds`:"
+                f" {negative_prompt_embeds}. Please make sure to only forward one of the two."
+            )
+
+        if prompt_embeds is not None and negative_prompt_embeds is not None:
+            if prompt_embeds.shape != negative_prompt_embeds.shape:
+                raise ValueError(
+                    "`prompt_embeds` and `negative_prompt_embeds` must have the same shape when passed directly, but"
+                    f" got: `prompt_embeds` {prompt_embeds.shape} != `negative_prompt_embeds`"
+                    f" {negative_prompt_embeds.shape}."
+                )
+
+        if ip_adapter_image is not None and ip_adapter_image_embeds is not None:
+            raise ValueError(
+                "Provide either `ip_adapter_image` or `ip_adapter_image_embeds`. Cannot leave both `ip_adapter_image` and `ip_adapter_image_embeds` defined."
+            )
+
+        if ip_adapter_image_embeds is not None:
+            if not isinstance(ip_adapter_image_embeds, list):
+                raise ValueError(
+                    f"`ip_adapter_image_embeds` has to be of type `list` but is {type(ip_adapter_image_embeds)}"
+                )
+            elif ip_adapter_image_embeds[0].ndim not in [3, 4]:
+                raise ValueError(
+                    f"`ip_adapter_image_embeds` has to be a list of 3D or 4D tensors but is {ip_adapter_image_embeds[0].ndim}D"
+                )
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_latents
+    def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
+        if isinstance(generator, list) and len(generator) != batch_size:
+            raise ValueError(
+                f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"
+                f" size of {batch_size}. Make sure the batch size matches the length of the generators."
+            )
+
+        if latents is None:
+            latents = randn_tensor(shape, generator=generator, device=device, dtype=dtype)
+        else:
+            latents = latents.to(device)
+
+        # scale the initial noise by the standard deviation required by the scheduler
+        latents = latents * self.scheduler.init_noise_sigma
+        return latents
+
+    def prepare_image_latents(
+        self, image, batch_size, num_images_per_prompt, dtype, device, do_classifier_free_guidance, generator=None
+    ):
+        if not isinstance(image, (torch.Tensor, PIL.Image.Image, list)):
+            raise ValueError(
+                f"`image` has to be of type `torch.Tensor`, `PIL.Image.Image` or list but is {type(image)}"
+            )
+
+        image = image.to(device=device, dtype=dtype)
+
+        batch_size = batch_size * num_images_per_prompt
+
+        if image.shape[1] == 4:
+            image_latents = image
+        else:
+            image_latents = retrieve_latents(self.vae.encode(image), sample_mode="argmax")
+
+        if batch_size > image_latents.shape[0] and batch_size % image_latents.shape[0] == 0:
+            # expand image_latents for batch_size
+            deprecation_message = (
+                f"You have passed {batch_size} text prompts (`prompt`), but only {image_latents.shape[0]} initial"
+                " images (`image`). Initial images are now duplicating to match the number of text prompts. Note"
+                " that this behavior is deprecated and will be removed in a version 1.0.0. Please make sure to update"
+                " your script to pass as many initial images as text prompts to suppress this warning."
+            )
+            deprecate("len(prompt) != len(image)", "1.0.0", deprecation_message, standard_warn=False)
+            additional_image_per_prompt = batch_size // image_latents.shape[0]
+            image_latents = torch.cat([image_latents] * additional_image_per_prompt, dim=0)
+        elif batch_size > image_latents.shape[0] and batch_size % image_latents.shape[0] != 0:
+            raise ValueError(
+                f"Cannot duplicate `image` of batch size {image_latents.shape[0]} to {batch_size} text prompts."
+            )
+        else:
+            image_latents = torch.cat([image_latents], dim=0)
+
+        if do_classifier_free_guidance:
+            uncond_image_latents = torch.zeros_like(image_latents)
+            image_latents = torch.cat([image_latents, image_latents, uncond_image_latents], dim=0)
+
+        return image_latents
+
+    @property
+    def guidance_scale(self):
+        return self._guidance_scale
+
+    @property
+    def image_guidance_scale(self):
+        return self._image_guidance_scale
+
+    @property
+    def num_timesteps(self):
+        return self._num_timesteps
+
+    # here `guidance_scale` is defined analog to the guidance weight `w` of equation (2)
+    # of the Imagen paper: https://arxiv.org/pdf/2205.11487.pdf . `guidance_scale = 1`
+    # corresponds to doing no classifier free guidance.
+    @property
+    def do_classifier_free_guidance(self):
+        return self.guidance_scale > 1.0 and self.image_guidance_scale >= 1.0
+
+    @classmethod
+    def from_pretrained_with_fidelity(cls, pretrained_model_path, fidelity_mlp_path=None, **kwargs):
+        """Load the pipeline with a fidelity MLP model."""
+        pipeline = cls.from_pretrained(pretrained_model_path, **kwargs)
+        
+        if fidelity_mlp_path is not None:
+            from fidelity_mlp import FidelityMLP
+            # Load the FidelityMLP
+            hidden_size = pipeline.text_encoder.config.hidden_size
+            if os.path.exists(fidelity_mlp_path):
+                pipeline.fidelity_mlp = FidelityMLP.from_pretrained(fidelity_mlp_path)
+            else:
+                # Create a new one if path doesn't exist
+                pipeline.fidelity_mlp = FidelityMLP(hidden_size)
+                logger.warning(
+                    f"Fidelity MLP not found at {fidelity_mlp_path}, initialized a new one with hidden size {hidden_size}"
+                )
+            
+            # Move to the same device as the pipeline
+            pipeline.fidelity_mlp.to(pipeline.device)
+        
+        return pipeline